Forward and inverse design of ambient and high pressure superconductors using DFT and deep learning

Over the past few decades, finding new superconductors with a high critical temperature T_c has been a challenging task due to computational and experimental costs. There have been efforts to discover superconductors both at ambient condition as well as very high external pressures. In this talk, we present 1) a large DFT database for both types of superconductors, 2) forward regression deep-learning models using the DFT dataset for fast screening of potential materials, 3) generative deep-learning model using diffusion approach to generate crystals with desired properties with unique structures and chemical compositions. Specifically, we used a crystal diffusion variational autoencoder (CDVAE) along with atomistic line graph neural network (ALIGNN) pretrained models and the Joint Automated Repository for Various Integrated Simulations (JARVIS) superconducting database of density functional theory (DFT) calculations to generate new superconductors with a high success rate. We started with a DFT dataset of ~1000 bulk superconducting materials to train the diffusion model. We used the model to generate 3000 new structures, which along with pre-trained ALIGNN screening results in 62 candidates. For the top candidate structures, we carried out further DFT calculations to validate our findings. We extended this approach to high pressure (0 - 500 GPa) hydride superconductors, where we searched the JARVIS database and literature for potential candidates and performed DFT for ~1000 additional hydride materials. After discovering several super hydride compounds with a T_c above MgB₂ (> 39 K), we utilized our same deep learning/generative approach and DFT workflow to discover new hydride-based structures outside of the initial training. Our approaches go beyond the typical funnel-like materials design approaches and allow for the inverse design of next-generation materials.